Most AI governance today is a probabilistic guess dressed in policy language. Oak & Sparrow built something different: Gatekeeper, a deterministic pre-execution governance kernel that sits between an AI agent and its desired real-world action. It evaluates the action against a formally encoded ontology of statutes, delivering a byte-deterministic GREEN, YELLOW, or RED verdict sealed in a SHA-256 artifact chain—no LLMs in the enforcement layer. The ontology validates cleanly under HermiT and Pellet reasoners, with refusal as the default state and execution requiring proof. Underpinning it all is rigorous math: an information-geometric framework based on the Fisher-Johosky metric, boasting a publicly verifiable forecasting record at 8.65 sigma across 156 pre-registered daily observations—3.65 sigma beyond CERN's Higgs discovery threshold.

AI researchers, governance experts, policymakers, and deep tech builders seeking robust, verifiable safeguards for AI deployment.

Join us for an architecture walkthrough, live demo, and open Q&A to explore Gatekeeper's inner workings and real-world implications.

11:00 AM - Doors open and welcome

11:15 AM - Architecture walkthrough

12:15 PM - Live demo

1:00 PM - Open Q&A

2:00 PM - Event close

This event is part of Deep Tech Week. Create a free profile at deep-tech-week.com/register to help hosts approve your registration and build your schedule for the week.

Location

Frontier Tower 🧑‍🚀

995 Market St, San Francisco, CA 94103, USA

Presented by

Deep Tech Week

The official calendar of Deep Tech Week.

To submit an event go to: www.deep-tech-week.com/host

Hosted By

AI